Two in vivo digestibility trials with sheep were conducted to identify the minimum period length of feeding a new diet to obtain reproducible values of nutritional variables onward and the minimum length of collection period as to obtain maximal precision for each variable. Trial 1 was conducted with ten Polwarth male sheep (34 ± 5 kg body weight (BW)) throughout three 21-day periods, in a completely randomized two-way crossover design. The animals were divided into two groups (Group A and B, n = 5 per group) which were fed ad libitum with a sequence of the following diets throughout the periods: Group A: hay - hay plus concentrate - hay; Group B: hay plus concentrate - hay - hay plus concentrate. The concentrate was included in a proportion of 0.33 of the total diet. The intake, and the faecal and urinary excretion were measured daily throughout the experiment. For evaluating rumen fermentation variables, in Trial 2 four Santa Inês male sheep (65 ± 5 kg BW) fitted with ruminal cannula were used. The animals were randomly divided into two groups (n = 2 per group), and the trial was conducted through four 21 days experimental period, in a three-way crossover design, using experimental diets and feeding management similar to Trial 1. The results indicated that, even though no clear or consistent steady-state condition was identified for rumen fermentation or urinary excretion variables, the adaptation period for measuring OM digestibility in in vivo trials with sheep fed ad libitum where the diet shifts from one of only hay to another containing concentrate, or vice-versa, should be at least 12 days long. Moreover, although no precision improvement was obtained by increasing the collection period above 1 day for measuring OM digestibility, the minimal length of collection period should be 4 days for measuring faecal excretion variables and 7 days for measuring urinary excretion variables.